Skip to main content

Documentation Index

Fetch the complete documentation index at: https://carat.im/docs/llms.txt

Use this file to discover all available pages before exploring further.

Apr 23, 2026
Introducing ‘ChatGPT Images 2.0’: A New Image Model
ChatGPT Images 2.0 Launch
Meet the image model that plans and reviews every detail before it draws.We’ve added ‘ChatGPT Images 2.0’, OpenAI’s latest image generation model, to Carat. Instead of rendering the moment it sees your prompt, it plans the composition, reviews its own output, and refines before delivery. Expect accurate multilingual text rendering and flexible aspect ratios built in.Available on Web and App today.
Apr 10, 2026
Introducing ‘Seedance 2.0’: A New Video Model
Seedance 2.0 Launch
Smoother motion. Richer stories. A brand new video generation model just landed on Carat.We’ve added ‘Seedance 2.0’, a new video generation model, to Carat. Create videos with natural motion, subtle facial expressions, and multi-scene storytelling from a single prompt. Every scene also comes with vivid, lifelike sound, so your videos feel truly immersive.
  • 💡 How to access: Select ‘Seedance 2.0’ as your video model to start creating.
Use the prompt generator below to craft better prompts and create the videos you want with ease.➡️ carat.im/apps/seedance_prompt_generatorAvailable on Web and App today.
Apr 1, 2026
Same Quality, Half the Cost: Google ‘Veo 3.1 Lite’ Is Here
Veo 3.1 Lite Launch
Google’s latest video model just landed on Carat—with the same performance at half the price.We’ve added ‘Veo 3.1 Lite’, Google’s newest video generation model. It delivers the same capabilities as Veo 3.1 Fast at less than half the cost. Generate videos from text or images, in 720p to 1080p, with both landscape (16:9) and portrait (9:16) formats. Choose from 4, 6, or 8-second durations to fit your needs.Available on Web and App today.
Mar 31, 2026
PixVerse v6: Multi-Shot Video at a Lower Cost
PixVerse v6 Launch
One prompt, multiple scenes. PixVerse v6 splits your story into shots automatically—at a fraction of the cost.We’ve added ‘PixVerse v6’, the latest multi-shot video model, to Carat. One prompt auto-splits into multiple scenes with consistent characters and scene-matched audio—no extra steps. Supports 1080p, up to 15 seconds, at a lower cost than other multi-shot models.Available on Web and App today.
Mar 11, 2026
Choose Your Agent: Carat 1.0, Lite, and Soul Are Here
Three Carat Agents
Power up for complex tasks, or go light and fast for simple requests. Now you can choose the right Carat agent for your needs.
  • ‘Carat 1.0 Lite’ for everyday conversations and light tasks
  • ‘Carat 1.0’ for tasks that require context understanding and judgment
  • ‘Carat 1.0 Soul’ for complex, in-depth requests with top-tier performance
Select your agent from the chat input panel to get started.Available on Web today and App version 2.241.0+.
Feb 27, 2026
Introducing ‘Nano Banana 2’: Google’s Latest Image Model
Nano Banana 2
Faster. Sharper. More precise. Google’s newest image generation model just landed on Carat—try it today.We’ve added ‘Nano Banana 2’, Google’s latest image generation model, to Carat. Generate images with unmatched speed, accuracy, and efficiency. It also delivers clean, precise text rendering, so every word in your image lands exactly as intended.Available on Web and App today.
Feb 13, 2026
Carat Video Model Now Generates Audio
Carat Video Model Upgrade
No more generating audio separately. The Carat Video (LTX) model now creates scene-matched sound right alongside your video.Every video you generate now comes with dialogue, sound effects, and ambient audio that fit the scene—all in one step. We’ve also bumped the resolution from 576p to 768p for sharper visuals, at the same 33-credit cost.
  • Lip sync is fully supported.
Jan 30, 2026
Pro Features Made Easy: Introducing the ‘Apps’ Tab
Apps Tab
Complex tasks that are hard to put into words? Now you can complete them effortlessly with a dedicated UI!We’ve brought together specialized tools like Beauty Model Studio and Character-to-Real Studio all in one place. Instead of typing out long prompts, simply upload your files and pick a template to get the exact results you want—faster and more accurately than ever.
  • 💡 How to access: Find the ‘Apps’ icon in the Web sidebar or between the Home and Discover tabs on the App.
Available on Web today and App version 2.234.0+.
Jan 12, 2026
Usage Management Update
Usage Management Update
Manage your precious Usage more thoroughly and transparently! We’ve separated your subscription plans, top-ups, and free credits so you can check balances and expiration dates at a glance.
  • 💡 Tip: Multiple top-ups are now displayed individually. Easily track which credits expire first and see your upcoming refill dates!
Available on Web today and App version 2.230.0+.
Jan 7, 2026
Now Available: Voice Input
Voice Input feature
Say goodbye to typing. Start your next project with a simple voice command and work with your AI teammate more naturally.Available on Web today and App version 2.229.0+.Jump to Chat from Library
Jump to Chat from Library feature
No more searching for lost chat rooms. From any Library detail page, just tap ‘Open in Chat’ to go straight to your conversation history.
Dec 22, 2025
Introducing Consistency mode
Consistency feature
We’ve added ‘Consistency’ to the Carat main page.
  • Register reference images for Characters, Products, Styles, or Backgrounds to visually guide your generations.
  • Fully supports NanoBanana, NanoBanana Pro, and ChatGPT Image 1.5.
    • 💡 Tip: Use NanoBanana Pro for the highest quality consistency.
  • Paid users enjoy unlimited uploads, while Free users get 1 slot.
Live on Web today and rolling out to App v2.227.0.
Dec 19, 2025
Chat. Build. Launch. 🚀Turn your words into fully functional apps.
  • Mini Apps are here: Build Mini apps just by chatting. No coding knowledge required.
  • Create with AI Tools: Use Carat’s Image, Video, and Music generation features to instantly craft unique content and assets for your Mini App.
    • Native integration of generative AI tools inside the Mini Apps (for your end-users) will be available in a future update.
Live now on both Web and App. Update to the latest version to start building.